Intradialytic hypotension
Kuipers, Johanna
IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite from it. Please check the document version below.
Document Version
Publisher's PDF, also known as Version of record
Publication date: 2019
Link to publication in University of Groningen/UMCG research database
Citation for published version (APA):
Kuipers, J. (2019). Intradialytic hypotension: Prevalence, defintions, impact on quality of life. Rijksuniversiteit Groningen.
Copyright
Other than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons).
Take-down policy
If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim.
Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons the number of authors shown on this cover page is limited to 10 maximum.
Prevalence, Definitions & Impact on Quality of Life
Further financial support for the printing of this thesis was kindly provided by: University of Groningen (Rijks Universiteit Groningen)
Research Institute GUIDE, University Medical Center Groningen Dutch Kidney Foundation (Nierstichting)
V&VN Dialyse&Nefrologie V&VN Research Professionals Vifor Pharma Nederland B.V.
Design and lay-out Fred de Jong Printing
GVO drukkers & vormgevers B.V., Ede, The Netherlands ISBN 978-94-034-1696-0
ISBN 978-94-034-1695-3 (e-book)
All rights reserved. No parts of this book may be reproduced or transmitted in any form or by any means whatsoever without express written permission from the author.
Prevalence, Definitions & Impact on Quality of Life
Proefschrift
ter verkrijging van de graad van doctor aan de Rijksuniversiteit Groningen
op gezag van de
rector magnificus prof. dr. E. Sterken en volgens besluit van het College voor Promoties.
De openbare verdediging zal plaatsvinden op maandag 17 juni 2019 om 14.30 uur
door Johanna Kuipers
Copromotores Dr. C.F.M. Franssen Dr. R. Westerhuis Dr. W. Paans Beoordelingscommissie Prof. dr. S.P. Berger Prof. dr. J.P. Kooman Prof. dr. T. Jaarsma Paranimfen Binie Kuipers Dr. Karin Ipema
Chapter 2. Variability of pre-, intra-, and post-dialytic
blood pressures in the course of a week: a study in Dutch and US chronic hemodialysis patients
Am J Kidney Dis (2013)
Chapter 3. Causes and consequences of interdialytic
weight gain
Kidney Blood Press Res (2016)
Chapter 4. Prevalence of dialysis hypotension. - A three-month,
prospective study of 3818 hemodialysis sessions in 124 hemodialysis patients
BMC Nephrology (2016)
Chapter 5. The prevalence of intra-dialytic hypotension in hemodialysis
patients on conventional hemodialysis. A systemic review
with meta-analysis
Accepted for publication in American Journal of Nephrology
Chapter 6. Association between Quality of Life and various aspects
of intradialytic hypotension including patient-reported intradialytic symptom score
Accepted for publication in BMC Nephrology
Chapter 7. Summary and General discussion
Nederlandse samenvatting Dankwoord
About the author List of publications 19 43 65 89 129 151 159 169 179 183
Chapter 1
Introduction
Introduction Dialysis
Dialysis is a life-saving treatment for people with end stage renal disease (ESRD). It was implemented on a large scale in patient care in the late 60`s and early 70`s of the 20th century century¹. Given the fact that no cure is available for ESRD, kidney transplantation is the best possible renal replacement therapy, however due to shortness
of donor organs and high comorbidity in patients with ESRD this is not always possible
². Dialysis is a form of renal function replacement treatment to remove the waste
products and excess fluid. In 2017 almost 6500 Dutch patients are being treated with
dialysis³. Approximately 14% of patients is treated with peritoneal dialysis and 86% with haemodialysis. Life expectancy in patients undergoing dialysis is relatively limited. Half of the patients that start dialysis between the ages of 45 and 65 years, die within 5 years⁴. The characteristics of patients undergoing dialysis have changed dramatically over the years. Where first relatively young patients with glomerulonephritis or polycystic kidney disease were treated, the patient group now includes older multi-morbid people with diabetes mellitus or hypertension as the cause of kidney failure⁵.
Dialysis seriously impairs the quality of life due to a high symptom burden. Most patients have a fluid restriction and a strict diet with low sodium and potassium. They have an increased cardiovascular risk and their general condition often deteriorates over time⁶. The haemodialysis treatment schedule and treatment itself also has a great impact on the daily lives of patients. The treatment is mostly thrice weekly for 4 hours and causes large changes in the composition of the blood over a short period of time. The removal of the excess fluid can lead to intradialytic hypotension which is often symptomatic and persists for some time after the treatment was completed ⁴.
Intradialytic hypotension
During haemodialysis, fluid is withdrawn from the circulation by the artificial kidney. At the same time the excess fluid moves from the interstitial tissue back into the circulation. The rate at which the fluid is withdrawn from the circulation (the ultrafiltration rate) is almost always higher than the rate at which the fluid moves from the interstitial tissue into the circulation (refill rate). This disbalance causes a decline in blood volume and often a decrease in blood pressure⁷. Cardiovascular compensating mechanisms such as venous and arterial vasoconstriction, redistribution of blood from peripheral and splanchnic vascular beds to the central blood compartment and increases in heart rate and cardiac contractility help to maintain blood pressure during hypovolemia⁸. The presence of cardiac dysfunction⁹, autonomous neuropathy¹⁰ and use
higher body mass index, lower pre-dialytic SBP and a relatively high dialysate temperature, may also play a role in the failure of these compensating mechanisms¹³-¹⁴.
Intradialytic hypotension (IDH) is considered one of the most frequent
complications of haemodialysis treatment and is associated with increased cardiovascular morbidity and mortality¹⁵. Various reviews report that up to 50% of haemodialysis sessions are complicated by IDH¹³-¹⁶-²². However, studies on the prevalence of IDH are relatively scarce²³-²⁶.
Over the years, dialysis techniques have improved and there is more awareness of strategies to prevent IDH, e.g. by lowering the dialysate temperature²⁷-²⁸ and monitoring and restricting of relative blood volume changes²⁹. At the same time, the average age of dialysis patients as well as the proportion of patients with significant co-morbidities such as diabetes mellitus and heart failure have increased³⁰-³¹.
A complicating factor in the analysis of IDH is that many different definitions of hypotension are used in the literature. These vary from liberal definitions that only require a minimum fall (e.g. 20 or 30 mmHg) in systolic blood pressure (SBP)³²-³⁴ to strict definitions that require the combination of a clinical event and a nursing intervention in addition to a minimum fall in blood pressure¹¹-³⁵-³⁶.
Variability in Blood Pressure and interdialytic weight gain
Patients on a thrice-weekly haemodialysis scheme are generally more fluid overloaded at the start of the first dialysis session of the week compared with the second and third session of the week. This is caused by the longer interdialytic interval before the first dialysis session of the week, which results in a higher interdialytic weight gain (IDWG). These variations in fluid status over the week may have major consequences for blood pressure dynamics before and during the individual dialysis sessions because extracellular volume is a major determinant of BP in dialysis patients ³⁷-³⁹. Thus it is conceivable that the more pronounced fluid overload at the start of the first HD session translates to higher pre-dialysis blood pressures compared with the subsequent haemodialysis sessions of the week. However, this was never studied.
At the same time, IDWG is increasingly being recognized as an indicator of nutritional status⁴⁰-⁴². Malnutrition is considered as a major complication among haemodialysis patients and can result in increased morbidity and mortality⁴³-⁴⁴. Several studies demonstrated that a greater IDWG is related to overall increase intake and is thus (directly) associated with improved nutritional status⁴¹-⁴²-⁴⁵.
Quality of Life
In the past 10-15 years, there has been an increasing awareness that patient survival might not be the one relevant outcome factor for patients who have a chronic disease. Quality of Live (QOL) has taken a more prominent place in research, and has been investigated in patients with chronic diseases, such as ESRD who depend on dialysis⁴⁶-⁴⁷. It has become increasingly important to not only provide the patients with the best medical and nursing care, but also to understand what significantly influences their QOL.
Several studies have investigated QOL in dialysis patients. Especially the physical components of QOL have been shown to be worse among dialysis patients in comparison with the general population⁴⁸, pre-dialysis CKD patients⁴⁹ and patients with other chronic diseases like congestive heart failure, diabetes, depression and even cancer⁵⁰. An improvement in QOL of dialysis patients was seen after renal transplantation: six months after transplantation, the mean health-related quality of life scores of almost all components of QOL had improved compared to pre-transplantation and remained
improved throughout the two years of follow up⁵¹-⁵².
IDH is often accompanied by symptoms such as nausea, dizziness, light-headedness, fatigue, and muscle cramps, affecting the daily lives of haemodialysis patients⁵³ and, therefore, likely influences QOL. Pathophysiology of IDH and the methods to avoid this complication have been extensively investigated⁵⁴-⁵⁵. Also the association between IDH and mortality has been studied by several groups, Flythe et.al. showed that an absolute nadir systolic blood pressure <90 mmHg was most potently associated with mortality⁵⁵. Caplin et.al. studied the burden and duration of haemodialysis-associated symptoms with a survey but did not study the association between symptoms and QOL⁵³.
As far as we know, there is no research on the association between intradialytic
symptoms and QOL.
The impact of dialysis on the lives of patients and their family members can be overwhelming⁵⁶. The nephrology nurses have long-term relationships with HD patients and often are familiar with the daily lives of patients. A better understanding of the factors that influence QOL of patients can help the nursing staff to support the patient in improving their QOL. This might lead to better motivated patients and better adherence to their treatment⁵⁷. Therefore, better knowledge on the association between QOL and haemodialysis treatment related factors like IDH and the impact on QOL can help to improve the QOL of patients.
Aim and outline of this thesis
The major aim of this thesis is to get a better insight in the multifactorial issues surrounding IDH and to contribute in finding a definition that can be used to best describe IDH also capturing the patient- experienced burden of IDH on QOL.
Patients on a trice weekly haemodialysis schedule have higher pre-dialysis weights and higher ultrafiltration rates at the first compared with the second and third dialysis session of the week. In chapter 2 we questioned whether these variations in excess
weight and ultrafiltration rate are associated with a consistent difference in pre-, intra-and post-dialysis blood pressure behaviour between the first intra-and the subsequent dialysis session of the week. In chapter 3 we aimed to identify the major determinants of a high
IDWG and its association with nutritional parameters. The major goal of the study described in chapter 4 was to assess the prevalence of IDH in our population and to
identify patient and treatment factors that are associated with its presence. In chapter 5 we describe a systemic literature review and meta-analysis on studies that investigated
the prevalence of IDH and provide an insight in the wide variation of definitions that are being used. In chapter 6 we studied whether the occurrence of IDH according to the
European Best Practice Guideline (EBPG) on haemodynamic instability as well as its 3 components, i.e. a fall in SBP of ≥20 mmHg, the occurrence of clinical events, and nursing interventions has an influence on the perception of QOL in HD patients. In Chapter 7 we
REFERENCES
1. Https://www.nierstichting.nl/over-de-nierstichting/missie/visiestrategie/ inspirator/.
2. Kievits F, Van Maanen H. Preëmptieve niertransplantatie onbenut. Ned Tijdschr
Geneeskd. 2009;153(C107).
3. Https://Www.nefrovisie.nl/renine/nefrodata.
4. Https://www.nierstichting.nl/professionals/factsheets/.
5. Jager,K.J., Ocak,G., Drechsler,C., et al. The EQUAL study: A european study in chronic kidney disease stage 4 patients. Nephrol Dial Transpl. 2012;27(suppl_3):iii27-iii31.
6. Janssen D, Spruit MA, Wouters E, Schols J. Daily symptom burden in end-stage chronic organ failure: A systematic review. Palliat Med. 2008;22(8):938-948. 7. Dasselaar JJ. Relative blood volume measurements during hemodialysis: Merits and
limitations. Rijksuniversiteit Groningen; 2007.
8. Ettema EM. Hemodynamic stability during hemodialysis: A role for vasopressin? Rijksuniversiteit Groningen; 2016.
9. van der Sande FM. The hemodynamic effect of different ultrafiltration rates in patients with cardiac failure and patients without cardiac failure: Comparison be tween isolated ultrafiltration and ultrafiltration with dialysis.Clin Nephrol. 1998-11;50(5):301-8.
10. Shafi Tariq T. Autonomic dysfunction as a mechanism of intradialytic blood pres sure instability. Semin Dial. 2017;30(6):537-544.
11. Kooman J, Basci A, Pizzarelli F, et al. EBPG guideline on haemodynamic instability.
Nephrol Dial Transplant. 2007;22 Suppl 2:ii22-ii44.
12. Chou JA. A brief review of intradialytic hypotension with a focus on survival.
Semin Dial. 2017;30(6):473-480.
13. Sands J, Usvyat LA, Sullivan T, et al. Intradialytic hypotension: Frequency, sources of variation and correlation with clinical outcome. Hemodial Int. 2014;18:415-422. 14. Selby NM, McIntyre CW. A systematic review of the clinical effects of reducing di
alysate fluid temperature. Nephrol Dial Transplant. 2006;21(7):1883-1898.
16. Davenport A. Can advances in hemodialysis machine technology prevent lytic hypotension? Semin Dial. 2009;22(3):231-236.
17. Dubin R, Owens C, Gasper W, Ganz P, Johansen K. Associations of endothelial function and arterial stiffness with intradialytic hypotension and hypertension.
modial Int. 2011;15(3):350-358.
18. Perazella MA. Pharmacologic options available to treat symptomatic intradialytic hypotension. Am J Kidney Dis. 2001;38(4 Suppl 4):S26-S36.
19. Passauer J, Büssemaker E, Gross P. Dialysis hypotension: Do we see light at the end of the tunnel? Nephrol Dial Transplant. 1998;13(12):3024-3029.
20. Daugirdas JT. Pathophysiology of dialysis hypotension: An update. Am J Kidney Dis. 2001;38(4, Supplement 4):S11-S17.
21. Daugirdas JT. Preventing and managing hypotension. Semin Dial. 1994;7(4):276-283.
22. Zucchelli P, Santoro A. Dialysis-induced hypotension: A fresh look at pathophysiol ogy. Blood Purif. 1993;11(2):85-98.
23. Tislér A, Akócsi K, Hárshegyi I, et al. Comparison of dialysis and clinical characteris tics of patients with frequent and occasional hemodialysis-associated hypotension.
Kidney Blood Press Res. 2002;25(2):97-102.
24. Converse RL, Jacobsen TN, Jost CM, et al. Paradoxical withdrawal of reflex vaso constriction as a cause of hemodialysis-induced hypotension. J Clin Invest. 1992;90(5):1657-1665.
25. Hossli S. Clinical management of intradialytic hypotension: Survey results. Nephrol
Nurs J. 2005;32(3):287-91.
26. Oliver MJ, Edwards LJ, Churchill DN. Impact of sodium and ultrafiltration profiling on hemodialysis-related symptoms. J Am Soc Nephrol. 2001;12(1):151-156. 27. Chesterton LJ, Selby NM, Burton JO, McIntyre CW. Cool dialysate reduces asymp
tomatic intradialytic hypotension and increases baroreflex variability. Hemodial
Int. 2009;13(2):189-196.
28. Jefferies HJ, Burton JO, McIntyre CW. Individualised dialysate temperature im proves intradialytic haemodynamics and abrogates haemodialysis-induced myo cardial stunning, without compromising tolerability. Blood Purif. 2011;32(1):63-68. 29. Franssen CFM, Dasselaar J, Sytsma P, Burgerhof JGM, Jong de PE, Huisman RM. Automatic feedback control of relative blood volume changes during hemodialysis improves blood pressure stability during and after dialysis. Hemodial Int. 2005;9(4):383-392.
30. Davenport A, Cox C, Thuraisingham R. Blood pressure control and symptomatic intradialytic hypotension in diabetic haemodialysis patients: A cross-sectional sur vey. Nephron.Clinical practice. 2008;109(2):c65-c71.
31. Mc Causland FR, Brunelli SM, Waikar S. Dialysis dose and intradialytic hypotension: Results from the HEMO study. Am J Nephrol. 2013;38(5):388-396.
32. Zimmerman CY, Ladefoged A, Søren SN. Can haemodialysis-induced hypotension be predicted? Nephron. 2002;92(3):582-588.
33. Schreiber MJ. Setting the stage. Am J Kidney Dis. 2001;38(4, Supplement 4):S1-S10. 34. Shimizu K, Kurosawa T, Sanjo T. Effect of hyperosmolality on vasopressin secretion in intradialytic hypotension: A mechanistic study. Am J Kidney Dis. 2008;52(2):294-304.
35. Prakash S, Garg AX, Heidenheim AP, House A. Midodrine appears to be safe and effective for dialysis-induced hypotension: A systematic review. Nephrol Dial
Transplant. 2004;19(10):2553-2558.
36. Bossola M, Laudisio A, Antocicco M, et al. Intradialytic hypotension is associated with dialytic age in patients on chronic hemodialysis. Ren Fail. 2013;35(9):1260-1263.
37. Hall JE. Pressure-volume regulation in hypertension. Kidney int. Supplement. 1996-6;55:35-41.
38. Koomans HA. The importance of plasma protein for blood volume and blood pressure homeostasis. Kidney Int. 1986-11;30(5):730-5.
39. Ventura JE. Volume sensitivity of blood pressure in end-stage renal disease.
Nephrol Dial Transplant. 1997-3;12(3):485-91.
40. Kalantar-Zadeh K. Fluid retention is associated with cardiovascular mortality in patients undergoing long-term hemodialysis. Circulation. 2009-2-10;119(5):671-9. 41. López-Gómez JM. Interdialytic weight gain as a marker of blood pressure,
nutri-tion, and survival in hemodialysis patients. Kidney Int. 2005-1(93):63-8.
42. Testa A. The other side of the coin: Interdialytic weight gain as an index of good nutrition. Am J Kidney Dis. 1998-5;31(5):830-4.
43. Araújo de IC. Nutritional parameters and mortality in incident hemodialysis pa-tients. J Ren Nutr. 2006-1;16(1):27-35.
44. Herselman M. Protein-energy malnutrition as a risk factor for increased
45. Sherman RA. Interdialytic weight gain and nutritional parameters in chronic hemodialysis patients. Am J Kidney Dis. 1995-4;25(4):579-83.
46. Merkus MP, Jager KJ, Dekker FW, Boeschoten EW, Stevens P, Krediet RT. Quality of life in patients on chronic dialysis: Self-assessment 3 months after the start of
treatment. Am J Kidney Dis. 1997;29(4):584-592. doi:
https://doi.org/10.1016/S0272-6386(97)90342-5.
47. Mingardi G, Cornalba L, Cortinovis E, Ruggiata R, Mosconi P, Apolone G. Health-re lated quality of life in dialysis patients. A report from an italian study using the SF-36 health survey. Nephrol Dial Transplant. 1999;14(6):1503.
48. Molsted S, Prescott L, Heaf J, Eidemak I. Assessment and clinical aspects of health-related quality of life in dialysis patients and patients with chronic kidney disease.
Nephron Clin Pract. 2007;106(1):c24-33.
49. Perlman RL, Finkelstein FO, Liu L, et al. Quality of life in chronic kidney disease (CKD): A cross-sectional analysis in the renal research institute-CKD study. Am J
Kidney Dis. 2005;45(4):658-666.
50. Mittal SK. Self-assessed physical and mental function of haemodialysis patients.
Nephrol Dial Transplant. 2001-7;16(7):1387-94.
51. Laupacis A, Keown P, Pus N, et al. A study of the quality of life and cost-utility of renal transplantation. Kidney Int. 1996;50(1):235-242.
52. Jofre R, López-Gómez JM, Moreno F, Sanz-Guajardo D, Valderrábano F. Changes in quality of life after renal transplantation. Am J Kidney Dis. 1998;32(1):93-100. 53. Caplin B, Kumar S, Davenport A. Patients' perspective of haemodialysis-associated
symptoms. Nephrol Dial Transplant. 2011;26(8):2656-2663.
54. Tisler A, Akocsi K, Borbas B, et al. The effect of frequent or occasional dialysis as-sociated hypotension on survival of patients on maintenance haemodialysis.
Neph-rol Dial Transplant. 2003;18(12):2601-2605.
55. Flythe JE, Xue H, Lynch KE, Curhan GC, Brunelli SM. Association of mortality risk with various definitions of intradialytic hypotension. J Am Soc Nephrol. 2014. 56. White Y. The biopsychosocial impact of end-stage renal disease: The experience
of dialysis patients and their partners. J Adv Nurs. 1999-12;30(6):1312-20.
57. Barnett T. Fluid compliance among patients having haemodialysis: Can an ed-ucational programme make a difference? J Adv Nurs. 2008-2;61(3):300-6.
Chapter 2
Variability of pre-, intra-, and post-dialytic blood pressures in
the course of a week: a study in Dutch and US chronic
hemodialysis patients
Johanna Kuipers Len A. Usvyat Jurjen K. Oosterhuis Judith J. Dasselaar Paul E. de Jong Ralf Westerhuis Jeffrey J. Sands Yuedong Wang Peter Kotanko Casper F.M. FranssenABSTRACT
Background: Patients with thrice-weekly hemodialysis have higher pre-dialysis weights and ultrafiltration rates at the first compared with subsequent dialysis sessions of the week. We hypothesized that these variations in weight and ultrafiltration rate are associated with a systematic difference in blood pressure.
Study Design: Observational study.
Setting and Participants: During three months we prospectively collected hemodynamic
data of 4007 hemodialysis sessions from 124 Dutch patients. A similar analysis was performed in 789 US patients comprising 6060 hemodialysis sessions.
Factor: First versus subsequent hemodialysis sessions of the week. Outcomes: Blood pressure.
Measurements: Blood pressure, weight, and ultrafiltration rate were analyzed separately
for the first, second, and third dialysis session of the week. Comparisons were made with linear mixed models.
Results: In Dutch patients pre-dialysis weight and ultrafiltration rate were significantly
greater at the first compared with subsequent hemodialysis sessions of the week (p<0.001). Pre-dialysis systolic and diastolic blood pressures were higher at the first than at subsequent sessions of the week (p<0.001). Pre-dialysis blood pressure differences persisted throughout the session: Systolic and diastolic blood pressures were on average 5.0 and 2.5 mmHg higher during the first compared to the third session of the week. Post-dialysis blood pressure followed a comparable pattern (p<0.001). Blood pressure differences between the first and subsequent days of the week persisted after adjustment for possible confounders. Results in the US cohort were materially identical despite differences in patient characteristics and treatment practice between the two cohorts.
Limitations: Dry weight was not assessed by objective methods.
Conclusion: Blood pressure of patients on a thrice-weekly dialysis schedule varies
systematically over the week. Pre-dialysis blood pressure is highest at the first hemodialysis session of the week, most likely due to greater interdialytic weight gain. Intra- and post-dialytic blood pressures are also highest at the first session of the week despite higher ultrafiltration rates.
INTRODUCTION
Patients on a thrice-weekly hemodialysis (HD) scheme are generally more fluid overloaded at the start of the first dialysis session of the week compared with the second and third session of the week. This is caused by the longer interdialytic interval before the first dialysis session of the week, which results in a higher interdialytic weight gain. These variations in fluid status over the week may have major consequences for blood pressure (BP) dynamics before and during the individual dialysis sessions.
Extracellular volume is a major determinant of BP in dialysis patients¹-³. Thus it is conceivable that the more pronounced fluid overload at the start of the first HD session might translate to higher pre-dialysis BPs compared with the subsequent HD sessions of the week. At the same time, higher ultrafiltration rates have to be prescribed at the first compared with the subsequent sessions of the week in order to end the HD session at target weight⁴. Since ultrafiltration rate is a major determinant of hemodynamic stability during HD⁵-⁷ variations in ultrafiltration rate may result in differences in BP behavior between the first and subsequent sessions of the week.
In this study we tested the hypothesis that differences in pre-dialysis volume status and ultrafiltration rate between the first and subsequent HD sessions of the week are associated with a systematic variation in pre-dialysis BP and intra-dialytic BP course. For this purpose, we prospectively collected the hemodynamic data of 4007 dialysis sessions from 124 Dutch patients and analyzed the course of BP, weight, ultrafiltration rate, and the incidence of dialysis hypotension separately for the first, second and third dialysis session of the week. To assess the generalizability of our results we performed a similar analysis in a cohort of 789 dialysis patients from six United States (US) dialysis centers that are part of the Renal Research Institute (RRI).
METHODS
Patients: Dutch study population
Adult (≥18 years) patients from the Dialysis Center Groningen and the dialysis unit of the University Medical Center Groningen were eligible for this study when they fulfilled the following criteria: 1. Maintenance bicarbonate HD for >3 months; 2. Three times a week 4 hours HD schedule. The only exclusion criterion was absence of informed consent. The study was performed in accordance with the principles of the Declaration of Helsinki.
Study protocol: Dutch study population
During 3 months, we prospectively collected the hemodynamic data of all HD sessions from participating patients. Each session was evaluated for pre- and post-HD weight, pre-, intra- and post-dialytic BP and heart rate (HR), ultrafiltration volume, and ultrafiltration rate. BP and HR were measured with an automated oscillometric monitor before HD, at 10, 30, 60, 120, and 180 minutes intra-HD, at the end of the HD session (240 minutes), and 10 minutes post-HD.
Patients were either dialyzed at Monday, Wednesday and Friday or at Tuesday, Thursday and Saturday. Monday and Tuesday were defined as the first, Wednesday and Thursday as the second, and Friday and Saturday as the third session of the week. All parameters were analyzed separately for the first, second and third session of the week. HD sessions during hospitalization were excluded from the analysis.
Prescriptions regarding target weight and antihypertensive medications were made by the nephrologist during weekly visits. Target weight was evaluated clinically (peripheral edema, signs of pulmonary congestion, intra- and extra-dialytic BP course, muscle cramps) and by (changes of) the cardiopulmonary radiological aspect. Excess weight at the start of HD was defined as the difference between pre-dialysis weight and target weight. The prescribed ultrafiltration volume was calculated by adding the estimated intra-dialytic fluid intake (usually 750 ml) to the excess weight. For all analyses, however, the exact ultrafiltration volume as ‘delivered’ by the dialysis apparatus was used. Ultrafiltration rate was calculated by dividing the ultrafiltration volume by dialysis session length and target weight. No saline was routinely administered to the patient during connection to the extracorporeal circuit. Individual treatment times were not changed in the course of the study.
Dialysis hypotension was defined as a decrease in systolic BP ≥30 mmHg in combination with systolic BP drop to <90 mmHg. We restricted this definition to BP since detailed per treatment information on symptoms and interventions were not available in the US cohort. Presence of residual renal function was indicated by a urinary volume of ≥500 ml per day.
Dialysis settings: Dutch Study population
All patients were dialyzed 3 times a week for 4 hours with low-flux polysulphone hollow-fiber dialysers, F8 or F10 (Fresenius Medical Care, Bad Homburg, Germany). Dialysis settings were identical for the first, second, and third session of the week. Blood flow ranged between 250 and 350 ml/min. Dialysate flow was 500 or 700 ml/min. Blood flow and dialysate flow were kept constant throughout the study period in the individual patient. All patients were dialyzed with a constant ultrafiltration rate. Dialysate temperature was 36.0 or 36.5 °C and was kept constant during the study period for the individual patient. Dialysate composition was as follows: sodium 139 mmol/l, potassium 1.0 to 3.0 mmol/l, calcium 1.5 mmol/l, magnesium 0.5 mmol/l, chloride 108 mmol/l, bicarbonate 34 mmol/l, acetate 3 mmol/l, and glucose 1.0 g/l. Patients received a light meal and 2 cups of coffee or tea during HD.
US study population
US patients were studied during the same 3-month calendar period using identical methodology as for the Dutch study population with some exceptions as outlined below. US patients were analyzed as part of standard operating procedures with BP recordings automatically collected by the computer system used in the RRI clinics. Consequently, BP and HR recordings were obtained not at pre-specified time points (as it was done in the Dutch study population) but as instructed by operational guidance which suggests measuring BP approximately every 30 minutes or when a patient experiences symptoms. The computer system that was used to collect intradialytic BPs was implemented about half way into the study period, so fewer treatments per patient were available in the US population.
Because US population was treated as part of standard clinic practice and patients were not enrolled into a study, it is possible that patients’ treatment times varied slightly from treatment to treatment. Furthermore, adjustments to blood and dialysate flow might have been made from treatment to treatment. Contrary to the Dutch population, patients were not dialyzed with constant ultrafiltration rate, in fact, in most cases, ultrafiltration rate was lowered in the last 15-30 minutes of the treatment. Dialysate temperature was between 35.0 and 37.0 °C. Dialysate composition differed: sodium 137 to 140 mmol/l, potassium 1.0 to 3.0 mmol/l, calcium 1.0 to 1.75 mmol/l, magnesium 0.8 to 1.0 mmol/l, chloride 101 to 107 mmol/l, bicarbonate 33 to 38 mmol/l, acetate 4 to 8 mmol/l, and glucose 1.0 g/l. Patients received no meals from the clinic but were allowed to bring their own meals to the dialysis unit.
Statistical analysis
test for categorical variables. Comparisons of weight, BP, ultrafiltration volume and rate between the first, second and third day of the week were made with linear mixed models with fixed and random effects accounting for inter- and intra-patient variability. In separate analyses, pre- and post-dialysis parameters were adjusted for factors that may affect BP: age, gender, dialysis vintage, diabetic status, body mass index, residual renal function, co-morbidity of ischemic heart disease, co-morbidity of congestive heart failure, dialysate composition (magnesium, calcium, and sodium concentrations), albumin-corrected plasma calcium level, plasma phosphorus level, and number of different cardiovascular drugs (angiotensin-converting enzyme inhibitor, angiotensin-receptor blocker, beta-blocker, calcium channel blocker) used by the patient. These data were collected at the start of the study period, except for calcium and phosphate levels, which were measured twice during the study period and averaged. For linear mixed models, treatment day, age, gender, dialysis vintage, diabetic status, body mass index, residual renal function, co-morbidity of ischemic heart disease and congestive heart failure, dialysate composition, albumin-corrected plasma calcium, plasma phosphorus level, and number of cardiac drugs were treated as fixed variables. Random variables were patient specific intercept and day when patient was treated. Linear mixed models were also used to adjust for a possible effect of pre-dialysis systolic BP on post-dialysis systolic BP⁶.
BPs for Dutch patients were measured at pre-specified time points whereas BPs for US patients were not measured at pre-specified time points. For a proper comparison of the Dutch and US populations, we used the percentage of dialysis session elapsed. To investigate the trend of BP and HR over percent of total treatment time, we computed means and standard deviations at pre-specified time points for Dutch patients. For US patients, we fitted cubic spline models and constructed confidence intervals for the mean functions. The R package ASSIST was used to fit cubic spline models where the smoothing parameter was selected by the generalized maximum likelihood method⁸.
RESULTS
Characteristics of the Dutch and US patient cohorts are shown in Table 1.
Patients: Dutch Study population
Hundred twenty-four patients participated in this study. Mean (±SD) hemoglobin and albumin levels were 10.7±1.2 g/dL and 39±3.2 g/L, respectively. eKt/V was 1.37±0.3 per session. HD access was an arterio-venous fistula or PTFE graft in 77% of patients and a tunneled central venous catheter in 23% of patients. A total of 4007 hemodialysis sessions were analysed. The average number of HD sessions analyzed per patient was 32.
Table 1.
Patient characteristics.
All values are presented as mean (standard deviation) unless otherwise indicated. Abbreviations: ADPKD, adult dominant polycystic kidney disease; CCB, calcium channel blocker; ACE-I, angiotensin converting enzyme inhibitor; ARB, angiotensin receptor blocker; NS, not significant. *Cardiovascular history: any history of ischemic heart disease, congestive heart failure, stroke or peripheral vascular disease.
Patients: US study population
Seven hundred and eighty nine patients were studied. US patients were younger, more often male, and had a longer dialysis vintage compared with the Dutch population (Table 1). The proportion of patients with diabetes (41% versus 27%; P<0.01) was
compared with the Dutch population (208±29 versus 236±12 minutes; P<0.01). Hemoglobin was slightly but significantly higher in the US population compared with the Dutch population (Table 1). eKt/V did not differ significantly from the Dutch population (Table 1). HD access was an arterio-venous fistula or PTFE graft in 83% of patients and a tunneled central venous catheter in 17% of patients. A total of 9560 hemodialysis sessions were analyzed while 6060 sessions were studied where all three treatment day data was available per patient. The average number of HD sessions analyzed per patient was 12.
Pre- and post-dialysis weight and ultrafiltration volume: Dutch Study population
As shown in Table 2a, pre-dialysis weight was 0.58 and 0.73 kg higher at the first compared with the second and third session of the week, respectively (p<0.001). Post-dialysis weight was 0.24 and 0.28 kg higher at the first compared with the second and third session of the week, respectively (p<0.001). The excess weight at the start of HD was significantly higher at the first compared with the second and third session of the week (both p<0.001). Total ultrafiltration volume and ultrafiltration rate were significantly (both p<0.001) higher during the first compared with the second and third session of the week (Table 2a). Differences in pre-dialysis weight, excess weight at the start of HD, and ultrafiltration volume and rate between the first and the subsequent days of the week remained significant after multivariable adjustment (Supplementary file S1).
Blood pressure and heart rate: Dutch Study population
As shown in Table 2a, pre-dialysis systolic and diastolic BP were significantly higher (p<0.001) and HR was significantly lower (p<0.001) at the first compared with the second and third session of the week. BP fell during each dialysis day but the pre-dialysis differences persisted throughout the dialysis session. At most time points during the treatment, systolic and diastolic BPs were significantly higher at the first compared with the other two dialysis sessions of the week (Figure 1 left). Post-HD systolic and diastolic BPs were significantly higher (p<0.001) and HR was significantly lower (p<0.01) after the first compared with the other two sessions of the week (Table 2a). Differences in pre-and post-dialysis BPs between the first pre-and subsequent days of the week remained significant after multivariable adjustment (Supplementary file S1).
Differences in pre-dialysis (Figure 2, left panel) and post-dialysis (Figure 3, left panel) systolic and diastolic BP and pre-dialysis HR between the first and the other HD sessions of the week were consistent throughout the 3-month study period. The same was observed for the intra-dialytic BPs (data not shown).
Frequency of dialysis sessions with a BP drop of ≥30 mmHg was more pronounced on the first treatment of the week; however, no significant difference in dialysis hypotension was observed (Table 2a).
Pre- and post-dialysis weight and ultrafiltration volume: US Study population
In agreement with the Dutch data, pre-dialysis weight was 0.44 and 0.53 kg higher at the first compared with the second and third session of the week, respectively (Table 2b; p<0.001). Post-dialysis weight was 0.25 and 0.32 kg higher at the first compared with the second and third session of the week, respectively (p<0.001). The excess weight at the start of HD was significantly higher at the first compared with the second and third session of the week (both p<0.001), which is the same as seen in the Dutch population. Similar findings for ultrafiltration volume and rate were observed in US as in the Dutch study population. No differences, however, were observed between second and third HD treatments of the week. Differences in pre-dialysis weight, excess weight at the start of HD, and ultrafiltration volume and rate between the first and subsequent days of the week remained significant after multivariable adjustment (Supplementary file S2).
Blood pressure and heart rate: US Study population
As in the Dutch population, US patients had higher pre-dialysis systolic and diastolic BPs at the first compared with the second and third session of the week (Table 2b). No significant differences were observed in pre-dialysis HR between the first and the second treatment, but significant differences between first and third treatment were observed (Table 2b). BP fell during each dialysis day but the pre-dialysis differences persisted throughout the entire dialysis session. At most time points during the treatment, systolic and diastolic BPs were significantly higher at the first compared with the other two sessions of the week (Figure 1 right). Differences in pre- and post-dialysis BPs between the first and subsequent days of the week remained significant after multivariable adjustment, except for the difference in pre-dialysis diastolic BP between the first and third session of the week which lost its significance (Supplementary file S2). As shown in Figure 2 and Figure 3 (right panels), for the US cohort differences in pre- and post-dialysis systolic and diastolic BPs between the first and the other session of the week were consistent throughout the 3-month study period but less consistent and less pronounced than in the Dutch population. Contrary to the Dutch patients, differences in HR between days of the week were not observed.
Contrary to the Dutch study population, frequency of systolic BP drops ≥30 mmHg was more pronounced on the third HD day. Less hypotensive episodes occurred on the first HD day compared to second (p<0.05) and third HD day of the week (NS) (Table 2b).
Table 2a.
Weight, ultrafiltration volume and ultrafiltration rate, blood pressure, and heart rate: Dutch study population.
All values are presented as mean (95% confidence interval) using linear mixed effect model (unadjusted data). Abbreviations: HD, hemodialysis; UF, ultrafiltration; bpm, beats per minute.
Table 2b.
Weight, ultrafiltration volume and ultrafiltration rate, blood pressure, and heart rate: US study population.
All values are presented as mean (95% confidence interval) using linear mixed effect model (unadjusted data). Abbreviations: HD, hemodialysis; UF, ultrafiltration; bpm, beats per minute. * Denotes p<0.05,
** denotes p<0.01, and *** denotes p<0.001.
Figure 1.
Courses of intra-dialytic systolic blood pressure, diastolic blood pressure, and heart rate. For the Dutch cohort on the left, each line joints mean values of the first, second and third hemodialysis sessions at pre-specified time points. For the US cohort on the right, each line represents cubic spline estimate of the mean function for the first, second and third hemodialysis sessions. Shaded area around each line represents 95% confidence intervals.
Figure 2.
Pre-dialysis systolic and diastolic blood pressure, and heart rate throughout the study period in the Dutch study population (left) and US study population (right). Each line represents the mean value of the 124 Dutch patients (left) and 789 US patients (right). The error bars represent the 95% confidence interval.
Figure 3.
Post-dialysis systolic and diastolic blood pressure, and heart rate throughout the study period in the Dutch study population (left) and US study population (right). Each line represents the mean value of the 124 Dutch patients (left) and 789 US patients (right). The error bars represent the 95% confidence interval.
Interaction between pre-dialysis and post-dialysis blood pressure
Since pre-dialysis systolic BP influences the BP course during the subsequent dialysis session⁶ we performed an additional analysis in which we used linear mixed model to adjust for the effect of pre-dialysis systolic BP on post-dialysis systolic BP in addition to the other adjustments. When adjusted for these factors, differences in post-dialysis systolic BP disappeared in the US population and there was still a significant difference between first and second day of the week in the Dutch population (Supplementary file S3).
DISCUSSION
In this study we addressed the question whether variations in pre-dialysis fluid status and ultrafiltration rate in patients on a thrice-weekly dialysis scheme translate to a systematic difference in BP behaviour between the first and subsequent dialysis session of the week. We found that pre-dialysis systolic and diastolic BPs were significantly higher at the first compared with the two other dialysis sessions of the week. A second, unexpected, finding was that intra- and post-dialytic BPs were also significantly higher during the first compared with the other dialysis sessions of the week despite higher ultrafiltration rates.
It is well established that fluid status is a major determinant of BP in dialysis patients¹-³. Since patients on a thrice-weekly HD scheme are more fluid overloaded after the longest interdialytic interval it is not surprising that pre-dialysis BP was significantly higher at the first compared with the other sessions of the week. Additionally, volume-independent mechanisms may contribute to the higher BP at the start of the first HD session of the week, e.g. by accumulation of uremic substances with vasopressor activity and/or a higher sympathetic tone after the long interdialytic interval.
The ultrafiltration rate is considered to be a major determinant of hemodynamic stability during HD⁵-⁶ and higher ultrafiltration rates are associated with increased cardiovascular morbidity and mortality⁹. The present study shows that intra- and post-dialysis BPs were higher at the first compared with the other post-dialysis sessions of the week despite higher ultrafiltration rates. Koomans et al and Wizeman et al have shown that the more fluid overloaded the patient, the smaller is the observed decrease in relative blood volume per unit of ultrafiltration volume⁸-⁹. This is explained by a higher refill rate from the interstitial tissues in a more fluid overloaded state⁸-⁹. As such, the higher degree of fluid overload at the first dialysis session of the week facilitates plasma refill and may explain that the BP fall during the first dialysis session did not exceed the
closer to target weight. This may also have contributed to lower intra- and post-dialytic BP at the second and third session of the week compared with the first session of the week.
In this study, interesting differences in patient characteristics and treatment practice between the US and the Dutch study population were observed. In particular, the proportion of patients with diabetes was higher in the US population whereas the proportion of patients with residual renal function was higher in the Dutch population. The average dialysis treatment time was significantly shorter in the US and this explains the higher ultrafiltration rate in the US compared with the Dutch population. Notably, differences in BP between the first and subsequent sessions of the week were comparable in the two populations despite these variations in patient and treatment characteristics.
The US cohort demonstrated significantly higher rates of dialysis hypotension (15.6%, 17.7% and 17.2% of treatments at the first, second, and third session of the week, respectively) than the Dutch cohort (7.4%, 6.6% and 7.2% of treatments at the first, second, and third session of the week, respectively) despite similar percentages of treatments with systolic BP declines ≥30 mmHg. The etiology of this difference remains unclear, however may relate to differences in ultrafiltration rate, dialysate temperature, the proportion of patients with diabetes, or other undetermined differences in patient characteristics or practice patterns between the US and Dutch centers.
Our findings may have relevance for the observation that the mortality rate in dialysis patients varies over the week and is influenced by the dialysis schedule¹²-¹⁵. A higher BP at the first dialysis day of the week may be one of the cardiovascular stressors that contribute to the higher cardiovascular event rate at the first dialysis day of the week. Our data support the claim that it is time to revisit the thrice-weekly conventional HD approach¹⁵. However, whether alternate-day hemodialysis reduces the differences in BP between the first and subsequent sessions of the week remains to be studied.
A limitation of this study is that we did not use objective methods to assess dry weight. Therefore, we cannot exclude that a proportion of patients was not at their true dry weight at the end of dialysis and this may have affected the BP level. However, this could not explain differences in BP between the different sessions of the week. A second limitation is the use of dialysis machine-measured BP which is subject to differing calibrations that may introduce bias. Bias in BP measurements could also be introduced by underlying vascular disease or prior access surgeries. This may be particularly rele-vant to the US population who had a higher percentage of diabetic patients. A third limitation is that we did not have detailed information on interventions for hemody-namic instability and treatment-to-treatment adaptations of dry weight that may affect BP behavior. Interventions during hemodialysis, however, would not affect the conclu-sion that pre-dialysis BP and weight are highest on the first treatment day of the week.
Finally, detailed information on (changes in) cardiovascular medication dosage, dialyz-ability, and long- versus short-acting status was not available.
In conclusion, this international multi-center study shows that pre-dialysis BP is highest at the first dialysis session of the week, probably due to more pronounced fluid overload. Despite significantly higher ultrafiltration rates, intra- and post-dialysis BP are also highest during the first session of the week compared with the subsequent dialysis sessions of the week.
ACKNOWLEDGEMENTS
Ada Schokkenbroek, Margreet in der Maur, and Crista Veldman are acknowledged for their assistance in data collection.
REFERENCES
1. Hall JE, Guyton AC, Brands MW. Pressure-volume regulation in hypertension.
Kidney Int. 1996;55(6):S35-41.
2. Koomans HA, Braam B. The importance of plasma protein for blood volume and blood pressure homeostasis. Kidney Int. 1986;30(5):730-735.
3. Ventura JE, Sposito M. Volume sensitivity of blood pressure in end-stage real disease. Nephrol Dial Transplant. 1997;12(3):485-491.
4. Dasselaar JJ, de Jong PE, Huisman RM, Franssen CFM. Influence of ultrafiltration volume on blood volume changes during hemodialysis as observed in day-of-the-week analysis of hemodialysis sessions. ASAIO J. 2007;53(4):479-484.
5. Daugirdas JT. Pathophysiology of dialysis hypotension: an update. Am J Kidney Dis. 2001;38(4 Suppl 4): S11-S17.
6. Dinesh K, Kunaparaju S, Cape K, Flythe JE, Feldman HI, Brunelli SM. A model of systolic blood pressure during the course of dialysis and clinical factors associated with various blood pressure behaviors. Am J Kidney Dis. 2011;58(5):794-803. 7. Davenport A. Can advances in hemodialysis machine technology prevent
intradialytic hypotension? Semin Dial. 2009;22(3):231-236.
8. Wang Y. Smoothing Splines: Methods and Applications. New York: Champman and Hall; 2011.
9. Flythe JE, Kimmel SE, Brunelli SM. Rapid fluid removal during dialysis is associated with cardiovascular morbidity and mortality. Kidney Int. 2011;79(2):250-257. 10. Koomans H, Geers A, Mees EJ. Plasma volume recovery after ultrafiltration in
patients with chronic renal failure. Kidney Int. 1984;26(6):848–854.
11. Wizemann V, Leibinger A, Mueller K, Nilson A. Influence of hydration state on plasma volume changes during ultrafiltration. Artif Organs 1995;19(5):416–419. 12. Foley RN, Gilbertson DT, Murray T, Collins AJ. Long interdialytic interval and
mortality among patients receiving hemodialysis. New Engl J Med. 2011;365(12):1099-1107.
13. Zhang H, Schaubel DE, Kalbfleisch JD, et al. Dialysis outcomes and analysis of practice patterns suggests the dialysis schedule affects the day-of-the week mortality. Kidney Int. 2012;81(11):1108-1115.
14. Prabhalvalkar SM, Mullan RN, Cunningham GC, Harron CJ. Association between dialysis schedule and mortality in maintenance hemodialysis. Kidney Int. 2012;82(9):1034-1035.
15. Liu J, Foley RN. Alternate-day dialysis may be needed for hemodialysis patients.
Kidney Int. 2012;81(11):1055-1057.
m ent ar y file S1 . gh t, u ltr af iltr ati on v ol ume an d r ate, b lood p ressu re, an d h ear t r ate, ad ju sted an al ysi s#: Du tch stu dy p op ul ati on . l v al ues ar e p resen ted as mean (9 5% con fid en ce in ter val ) u sin g l in ear mi xed ef fect mod el . ll par ameter s ex cep t p re-d ial ysi s an d post-d ial ysi s w ei gh t w er e ad ju sted for age, gen de r, di al ysi s v in tage, d iab eti c statu s, bod y mass in dex , r esi du al ren al fu ncti on , co-mor bi di ty of isch emi c t d isease, co-mor bi di ty of con gesti ve hear t f ai lu re , d ial ysate comp osi tion (magn esi um, cal ciu m, an d sod iu m con cen tr ati on s), al bu mi n-cor rected p lasma cal ciu m lev el , p lasma ph osp hor us el , an d th e n umb er of d iff er en t car di ov ascu lar d ru gs. P re-d ial ysi s an d p ost-d ial ysi s w ei gh ts w er e ad ju sted for al l th ese f actor s ex cep t b od y mass i nd ex . br ev iati on s: HD, h emod ial ysi s; UF , u ltr af iltr ati on ; b pm, b eats p er mi nu te . ote s p <0 .0 5, ** d en otes p <0 .0 1, an d *** d en ote s p <0 .0 01 .
ar y file S2 . af iltr ati on v ol ume an d r ate, b lood p ressu re, an d h ear t r ate, ad ju sted an al ysi s#: US stu dy p op ul ati on . ted as mean (9 5% con fid en ce in ter val ) u sin g l in ear mi xed ef fect mod el . cep t p re-d ial ysi s an d post-d ial ysi s w ei gh t w er e ad ju sted for age, gen de r, di al ysi s v in tage, d iab eti c statu s, bod y mass in dex , r esi du al ren al fu ncti on , co-mor bi di ty of isch emi c bi di ty of con gesti ve hear t f ai lu re , d ial ysate comp osi tion (magn esi um, cal ciu m, an d sod iu m con cen tr ati on s), al bu mi n-cor rected p lasma cal ciu m lev el , p lasma ph osp hor us umb er of d iff er en t car di ov ascu lar d ru gs. P re-d ial ysi s an d p ost-d ial ysi s w ei gh ts w er e ad ju sted for al l th ese f actor s ex cep t b od y mass i nd ex . emod ial ysi s; UF , u ltr af iltr ati on ; b pm, b eats p er mi nu te . 5, ** d en otes p <0 .0 1, an d *** d en ote s p <0 .0 01 .
2
m ent ar y file S3 . ial ysi s b lood p ressu re i n th e Du tch an d US p op ul ati on , ad ju sted for p re-d ial ysi s sy stol ic b lood p ressu re#. dj us ted fo r a ge, gender, vint ag e, dia bet ic st at us , BM I, res idua l rena l f unc tio n, co m orbidit y o f is chem ic hea rt dis ea se, co m orbidit y o f c ong es tiv e hea rt fa ilure, dia lys at e co m po sition (m ag nesium , ciu m, an d sod iu m), p lasma cal ciu m, p lasma p hosp hor us, n umb er of car di ov ascu lar d ru gs, an d p re-d ial ysi s sy stol ic b lood p ressu re. otes p <0 .0 1
Chapter 3
Causes and consequences of interdialytic weight gain
*Johanna Kuipers and *Karin J.R. Ipema Ralf Westerhuis
Carlo A.J.M. Gaillard Cees P. van der Schans Wim P. Krijnen Casper F.M. Franssen *Contributed equally
ABSTRACT
Background: Higher interdialytic weight gain (IDWG) is associated with higher predialysis
blood pressure and increased mortality. IDWG is also increasingly being recognized as an indicator of nutritional status. We studied in detail the associations of various patient factors and nutritional parameters with IDWG.
Methods: We collected data during one week for IDWG and hemodynamic parameters
in 138 prevalent adult haemodialysis patients on a thrice-weekly haemodialysis schedule. A multivariate linear regression analysis was employed to identify factors that are associated with IDWG.
Results: The mean (±SD) age was 62.5 (±18.2) years, 36% were female, 36% had diuresis,
and 23% had diabetes. Patients in the highest IDWG tertile were significantly younger, more frequently male, and had a significantly higher subjective global assessment score (SGA). A higher IDWG as a percentage of body weight (%IDWG) was associated with a younger age, greater height and weight, absence of diuresis, and lower postdialysis plasma sodium levels. The model with these five parameters explained 37% of the variance of %IDWG. Predialysis, intradialysis, and postdialysis diastolic blood pressure was significantly higher in the highest tertile of IDWG.
Conclusion: The most important associations of %IDWG are age, height, weight, diuresis,
and postdialysis sodium. Patients with the highest IDWG have significantly higher diastolic blood pressures.
BACKGROUND
Interdialytic weight gain (IDWG) is the result of salt and water intake between two haemodialysis sessions. IDWG is used as a parameter for fluid intake while taking the daily urine output into account¹-² A higher IDWG is associated with higher predialysis blood pressure²-³, greater intradialytic reductions in blood pressure as a result of higher ultrafiltration rates⁴, and increased mortality⁵-⁷.
At the same time, IDWG is increasingly being recognized as an indicator of nutritional status⁸-⁹-¹⁰-¹¹. Malnutrition is considered as a major complication among haemodialysis patients and can result in increased morbidity and mortality¹²-¹³ Causes of malnutrition in dialysis patients are multi-factorial and include reduced appetite and food intake¹²-¹⁴-¹⁵, protein-energy wasting as a result of chronic inflammation¹⁶, and reduced physical activity¹⁷. Several studies demonstrated that a greater IDWG is directly associated with improved nutritional status²-¹⁰-¹¹. Usvyat et al. recently showed that IDWG began to decline a year before death indicating that a decrease in IDWG has short-term adverse prognostic significance¹⁸. Thus, on the one hand, higher IDWG is associated with adverse effects such as higher blood pressure, however, on the other hand, higher IDWG may be associated with favourable effects such as better nutritional status.
The goal of this study was to identify the most important associations of a high IDWG in an effort to disentangle its ambiguous associations. To achieve this, we meticulously examined a cohort of 138 patients on a thrice-weekly haemodialysis schedule.
METHODS
Participants and Study design
We retrospectively collected data from 138 haemodialysis patients scheduled for thrice weekly haemodialysis who were older than 18 years and had been undergoing haemodialysis treatment for at least three months. Since IDWG tends to fall before death¹⁸ and this may confound the relationship between IDWG and nutritional status in patients with a short life expectancy, we excluded patients who died within 6 months after collection of the data. We used data of IDWG, various nutritional parameters, and hemodynamic measurements during one week from the patients' records in November 2012. The study was performed in accordance with the principals of the Declaration of Helsinki and guidelines for Good Clinical Practice.
Dialysis regimens and Dietary consultation
Dialysis treatment consisted of conventional haemodialysis or home haemodialysis thrice weekly for four to five hours with blood flows and dialysate flows of 250-350 ml/min and 500-700 ml/min, respectively. All patients were dialyzed with low-flux polysulphone dialyzers and a constant dialysate conductivity of 13.9 mS/cm. The dialysate composition was as follows: sodium 139 mmol/l, potassium 1.0 or 2.0 mmol/l, calcium 1.5 mmol/l, magnesium 0,5 mmol/l, chloride 108 mmol/l, bicarbonate 34 mmol/l, acetate 3 mmol/l, glucose 1.0 g/l. Low-molecular-weight heparin was used as an anticoagulant.
Dry weight was evaluated clinically (peripheral oedema, signs of pulmonary congestion, intra- and interdialytic blood pressure course, muscle cramps) in combination with the predialysis cardiothoracic ratio on a chest X-ray as a surrogate marker of hydration status.
All patients had regular contact with the dietician every four to six weeks according to usual clinical practice. During these visits, the nutritional status was evaluated, and changes in weight, laboratory results, and appetite were monitored.
Measurements
For all of the patients, we collected demographic data including age, gender, level of education, and patient characteristics such as dialysis vintage, weight, and height. Body mass index (BMI) was calculated as: postdialysis weight (kg)/length (m)². Cardiovascular history was defined as any history of ischemic heart disease, congestive heart failure, stroke or peripheral vascular disease, and hypertension. Residual renal function was defined as diuresis ≥200 ml/day. Equilibrated Kt/V was calculated from pre-and postdialysis plasma urea concentration according to the second generation logarithmic Daurgirdas equation¹⁹.
The nutritional status of the patients was assessed with various parameters: the seven-point subjective global assessment (SGA), serum albumin, dry body weight, body height, BMI, and protein catabolic rate (PCR). The SGA has been described and validated in dialysis patients in the Netherlands Cooperative Study on the Adequacy of Dialysis²⁰. A score of ‘1’ indicates severe protein energy wasting, and a score of ‘7’ indicates a normal nutritional status. Blood samples were collected in heparin-coated tubes from the arterial line at the initiation and at the end of the first haemodialysis session of the study week in order to determine sodium and albumin levels. Plasma sodium was measured with the indirect method of ion-selective electrode on a Roche Modular (Hitachi, Tokyo, Japan).
IDWG was calculated as predialysis weight minus the postdialysis weight of the previous haemodialysis session. Since body weight may influence nutritional and fluid intake, the results are also shown for IDWG as a percentage of dry body weight (%IDWG)⁹. The ultrafiltration rate was calculated by dividing the ultrafiltration volume (ml) by the length of time of the dialysis session (hours) and target dry weight (kg). Blood pressure was measured with an automatic oscillometric monitor that is incorporated in the haemodialysis apparatus. The results of IDWG, ultrafiltration volume and rate, and blood pressure for the three haemodialysis sessions in the study week were averaged.
Statistical Analyses
Data are reported as mean±SD for continuous variables with normal distributions and numbers (percent) for categorical data. Demographic characteristics, laboratory data, and blood pressures were categorized into tertiles of IDWG and %IDWG. Differences between tertiles were analysed with ANOVA followed by Tukey’s honest post hoc test. For categorical data, the Pearson Chi-Square test and the Generalized Cochran Mantel-Haenszel Test were used.
A multivariate linear regression analysis was utilized to identify patient factors including various nutritional parameters that were associated with IDWG and/or %IDWG. IDWG or %IDWG was entered as a response variable. The following possible explanatory variables were entered into the model: age, gender, weight, height, Kt/V, dialysis vintage, diuresis, diabetes, SGA, nPCR, serum albumin, and predialysis and postdialysis plasma sodium concentration (Figure 1). Next, to identify variables significantly contributing to IDWG, the Bayesian Information Criterion (BIC) for model selection was used²¹. Statistical analyses were performed with SPSS version 20 (SPSS inc., IBM company, USA) and statistical programming language R (R Development Core Team)²². Two-tailed P-values <0.05 were considered statistically significant.
Figure 1.
Table 1.
Patient characteristics for the total group and according to tertiles of absolute IDWG.
Abbreviations: SGA: subjective global assessment, (n)PCR: (normalized) protein catabolic rate, IDWG: Interdialytic weight gain, UF: ultrafiltration.
P values: differences in means between the 3 groups tested by ANOVA.
RESULTS
Patient characteristics
Patient characteristics are depicted in Table 1. The mean (±SD) age was 62.5 (±18.2) years, 36% were female, 36% had diuresis, and 23% had diabetes. Patients in the highest IDWG tertile were significantly younger (P=0.004), more frequently male (P=0.001), taller (P<0.0001), heavier (P=0.009), and had a significantly higher SGA (P=0.004) compared with patients in the other tertiles (Table 1). Similar results were obtained for %IDWG (data not shown).
IDWG and possible explanatory variables.
In the multivariate linear regression model with optimizing BIC, the response variable IDWG was significantly associated with the explanatory variables height, age, the presence of residual diuresis, and postdialysis sodium levels. The model incorporating these four variables explained 35% of the variance of absolute IDWG (Table 2). The response variable %IDWG was significantly associated with the presence of residual diuresis, age, weight, height, and post-dialysis sodium levels. The model with these five variables explained 37% of the variance of the %IDWG (Table 3). Height was positively associated with absolute IDWG and %IDWG. Weight was positively associated with %IDWG. Age had a negative effect on IDWG whereby one year of older age resulted in a decrease of 0.016 kg and 0.023% in absolute IDWG and %IDWG, respectively. The presence of residual diuresis was associated with a significantly lower IDWG and %IDWG. Postdialysis sodium levels had a negative association with both IDWG and %IDWG: higher postdialysis sodium levels were associated with lower IDWG and %IDWG. Since this was an unexpected finding, we analysed the course of pre- to postdialysis plasma sodium concentration per tertile (Supplementary file 1). Patients in the middle and highest IDWG tertiles had a lower plasma sodium concentration, both pre- and postdialysis, compared with patients in the lowest IDWG tertile (Table 1). However, differences between the IDWG tertiles were only significant for postdialysis sodium concentration (Table 1). The other tested dependent variables (Kt/V, dialysis vintage, diabetes, SGA, serum albumin, and predialysis plasma sodium level) did not significantly contribute to explaining the variance of absolute IDWG or %IDWG.
Table 2.
Multivariate linear regression analysis with model building strategy Bayesian Information Criterion (BIC) – factors that are associated with absolute IDWG.
IDWG was entered as a response variable, the other parameters as explanatory variables. The variance of absolute IDWG is explained for 35% by the explanatory variables. Abbreviations: SE: standard error, CI: confidence interval.
Table 3.
Multivariate linear regression analysis with model building strategy Bayesian Information Criterion (BIC) – factors that are associated with %IDWG.
Relative IDWG was entered as a response variable, and the other parameters as explanatory variables. The variance of %IDWG is explained for 37% by the explanatory variables.
Abbreviations: SE: standard error, CI: confidence interval.
Effect of gender and age
Table 1 shows that patients with the highest IDWG (tertile 3) were younger and more frequently male. As demonstrated in Figure 2a, younger males (median age ≤ 65 years (yr)) indeed had a significantly higher IDWG compared with younger females (median age ≤ 69.5 yr, P=0.002), older females (median age > 69.5 yr, P=0.000), and older males (median age > 65 yr, P=0.008). For %IDWG, there was only a significant difference between younger males (median age ≤ 65 yr) and older females (median age > 69.5 yr, P=0.030) (figure 2b).
Figure 2.
The combined effect of age and gender on absolute IDWG (upper panel) and %IDWG (lower panel).
IDWG and blood pressure
Pre-, intra- and postdialysis systolic blood pressure did not vary significantly between tertiles of IDWG. Predialysis, intradialysis, and postdialysis diastolic blood pressure (DBP) was significantly higher in the highest IDWG tertile compared with the lowest tertile (Table 4, Figures 3a and 3b). For %IDWG, predialysis and intradialysis DBP was significantly higher in the highest %IDWG tertile compared with the lowest tertile (Table 5, Figure 3c and 3d).
Table 4.
Differences in blood pressures between tertiles of absolute IDWG.
Abbreviations: IDWG: interdialytic weight gain, CI: confidence interval; SBP: systolic blood pressure. DBP: diastolic blood pressure.
#P value denotes differences between the tertiles with ANOVA. Differences between the groups were analysed with a post-hoc Tukey Honest test.
Table 5.
Differences in blood pressures within tertiles of %IDWG.
Abbreviations: IDWG: interdialytic weight gain, CI: confidence interval; SBP: systolic blood pressure; DBP: diastolic blood pressure.
#P value denotes differences between the tertiles with ANOVA. Differences between the groups were analysed with a post-hoc Tukey Honest test.
Figure 3.
Differences in systolic (left panel) and diastolic (right panel) blood pressures between absolute IDWG (upper panel) and %IDWG tertiles (lower panel).
